Controller Tuning

The performance of feedback controllers depends on the values of their chosen

parameters. If these parameters are properly chosen, they offer the highest flexibility
to achieving the desired controlled response and stability. The process of choosing
these parameters is known as controller tuning.
Also, controller tuning can be defined as an optimization process that involves a
performance criterion to the form of the controller response and to the error between
the process variable and the set point.

Controller Tuning
 Tuning is the adjustment of the controller parameters to obtain a specified closed
loop response. Tuning of controller is the process of setting controller gains to
achieve desired performance.
 After a Control system is installed the controller settings must usually be
adjusted until the control system performance is considered to be satisfactory.
 The goal of the tuning is to determine the gains that optimize system response.
High gain increase responsiveness but also move the system closer to instability.
 Low gain improve stability but the system becomes sluggish.

Design and Tuning of a Control System

1) What type of feedback controller should be used to control a given process?
Some engineer –select PI and other select PID. In all cases, the selectionof Kc
and τI have and important effect on the response of the controlled process.
2) How do we select the best values for the adjustable parameters of the feedback
controller?
This is known as the controller tuning problem. We need to have a quantitative
measure in order to compare the alternatives and select the best type of controller
and the best values of its parameters.
3) What performance criteian we use for the section and the tuning the controller?
There are a avrity of the performance criteria we would we, such as:
1) Keep the max deviation (Error) as small as possible.
2) Achive short settling time.
3) Minimize the integral of errors untill the process has settled to its setpoint,
and so on.
4) Keep low value of Decay ratio.

Process Control 42 Fourth Class

Dr. Zaidoon M. Shakor
Simple Performance Criteria
1) Steady-state performance criteria:
The principal steady-state performance criterian usually is zero. Such a steady-
state error is called an offset. The P control cannot achieve S.S. error, while a PI can.
2) Dynamic Response performance criteria:
a-Criteria that use only a few points of the response they are simpler but only
approximate.
b- Criteria that use the entire closed –loop response from t=0 until t=very large. There
are more precise but also more cumbersome to use.
For the Best criteria is the decay ratio, and the
optimum value of DR is.
𝐶 1
=
𝐴 4

It is reasonable tradeoff between a fast rise time and a

reasonable settling time.
This criteria is usually known as one-quarter decay ratio.

Tuning Relations Based on Integral Error Criteria

• Controller tuning relations have been developed that optimize the closed-loop
response for a simple process model and a specified disturbance or set-point
change.
• The optimum settings minimize an integral error criterion.
• Three popular integral error criteria are:
The main three methods of the integral error performance criteria used in terms
of:
 Integral of the absolute value of the error (IAE)
∞
𝐼𝐴𝐸 = ∫0 |𝑒(𝑡)| 𝑑𝑡 ….. (1)
where the error signal e(t) is the difference between the set point and the
measurement.
 Integrated Square Error (ISE)
This error uses the square of the error, thereby penalizing large errors more than
small errors. This gives more conservative response (faster return to set point).


ISE   e dt
2 ….. (2)
0

 Integrated Time-Weighted Absolute Error (ITAE)

Process Control 43 Fourth Class

Dr. Zaidoon M. Shakor
 This criterion is based on the integral of the absolute value of the error multiplied
by time. It results in errors existing over time being penalized even though may be
small, which results in a more heavily damped response.


ITAE   t e dt
….. (3)
0

If the performance indices increases, control system can perform poorly and
even become unstable. So it needs to tune the controller parameters to achieve good
control performance with the proper choice of tuning constants.

Example: Select the gain of a proportional controller (Kc) using the one-quarter decay
ratio criterion (DR=0.25).
The process is described by: Gp(s) = 10/[(s+2)(2s+1)] ,
Assume Gm(s) = Gv= 1.
Answer:
𝐺𝑃 𝐺𝑣 𝐺𝑐
𝑌(𝑠) =
1 + 𝐺𝑃 𝐺𝑣 𝐺𝑐 𝐺𝑚

2ψ

0.21083

=6.93

Process Control 44 Fourth Class

Dr. Zaidoon M. Shakor
Q2: Find the gain of a proportional controller that produces a closed-loop response for
a second-order system with decay ratio equal to 1/4 . The process is described by:-
1
GP (s) = 2 , Gm (s) = Gf (s) = 1
s + 3s + 1
Answer:
𝐺𝑃 𝐺𝑣 𝐺𝑐
𝑌(𝑠) =
1 + 𝐺𝑃 𝐺𝑣 𝐺𝑐 𝐺𝑚
1
𝐾𝑐 2
𝑌(𝑠) = s + 3s + 1
1
1 + 𝐾𝑐 2
s + 3s + 1
𝐾𝑐
2
𝑌(𝑠) = 2 s + 3s + 1
s + 3s + 1 + 𝐾𝑐
s 2 + 3s + 1
𝐾𝑐
𝑌(𝑠) = 2
s + 3s + 1 + 𝐾𝑐
𝐾𝑐
1 + 𝐾𝑐
𝑌(𝑠) = 2
s 3
+ s+1
1 + 𝐾𝑐 1 + 𝐾𝑐
1
𝜏=
√1 + 𝐾𝑐
3
2𝜉𝜏 =
√1 + 𝐾𝑐
3
2𝜉𝜏 =
1 + 𝐾𝑐
3√1 + 𝐾𝑐 3
𝜉= =
2(1 + 𝐾𝑐 ) 2√1 + 𝐾𝑐
For DR=1/4 the 𝜉 = 0.2154
3
0.2154 =
2√1 + 𝐾𝑐
√1 + 𝐾𝑐 = 6.9621
𝐾𝑐 =47.4718

Process Control 45 Fourth Class

Dr. Zaidoon M. Shakor
Classical methods for controller tuning
The following two methods are preferred here those of Cohen-Coon method also
known as process reaction curve derived from open-loop systems. The other is Ziegler-
Nichols method based on frequency response analysis (bode diagram) or continuous
cycling method that it derived mainly from closed-loop systems.

1) Cohen-Coon Method
It is an empirical rule which have been proven is practice, it is known as Cohen and
Coon method.

d(s)
Gd
1
C(s) 
s
R Y(s)
GC Gv Gp

Ym(s) Y (s)
Gm

The Cohen - Coon method applies only to open-loop processes that are inherently
stable.
Cohen and Coon suggest method to first model process in the open loop and then
picking the appropriate control parameters. Noted that most responses have a
“sigmoidal” response to step change.
The procedure of this method is to open the control system by disconnection the
controller from the final control element, and introduce a step change of magnitude A
in the variable L which actuates the final control element.
Then find the T.F. between Ym and C as
𝑌𝑚 (𝑠)
𝐺𝑃𝑅𝐶 = = 𝐺𝑓 (𝑠) 𝐺𝑝 (𝑠) 𝐺𝑚 (𝑠) ….. (4)
𝐶(𝑠)
Cohen-Coon used process reaction curve, that it is a response of the process to a step
change in the manipulated variable. Cohen and Coon observed that the response of
most processing units to a step change in input variable can be adequately
approximated by the response of first order system with dead time, and the transfer
function is:
𝑌𝑚 (𝑠) 𝐾𝑒 −𝑡𝑑𝑠
𝐺𝑃𝑅𝐶 = ≈ ….. (5)
𝐶(𝑠) 𝜏𝑠+1

Process Control 46 Fourth Class

Dr. Zaidoon M. Shakor
The values of K, τ and td are calculated from the process reaction curve which is shown
in Fig. (1). A tangent is drawn to the curve at the point of maximum rate or ascent, and
then td is the intercept of this tangent with x-axis, and is defined as the time elapsed
until the system responds.
𝐵 𝑂𝑢𝑡𝑝𝑢𝑡 (𝑎𝑡 𝑠𝑡𝑒𝑎𝑑𝑦 𝑠𝑡𝑎𝑡𝑒)
𝐾= = ….. (6)
𝐴 𝐼𝑛𝑝𝑢𝑡 (𝑎𝑡 𝑠𝑡𝑒𝑎𝑑𝑦 𝑠𝑡𝑎𝑡𝑒)
𝐵 𝑂𝑢𝑡𝑝𝑢𝑡 (𝑎𝑡 𝑠𝑡𝑒𝑎𝑑𝑦 𝑠𝑡𝑎𝑡𝑒)
𝜏= = ….. (7)
𝑆 𝑆𝑙𝑜𝑝𝑒
where S is the slope of the sigmoidal response at the point of inflection, td = time
elapsed until the system responded.
Cohen and coon used the approximated model and estimated K, τ, τd .

1) For Proportional controller:

1 𝜏 𝑡𝑑
𝐾𝐶 = (1 + ) ….. (8)
𝐾 𝑡𝑑 3𝜏
2) For Proportional-Integral controller:
1 𝜏 𝑡𝑑
𝐾𝐶 = (0.9 + ) ….. (9)
𝐾 𝑡𝑑 12𝜏
𝑡
30+3 𝑑
𝜏𝐼 = 𝑡𝑑 𝑡
𝜏
….. (10)
9+20 𝑑
𝜏
3) For Proportional-Derivative controller:
1 𝜏 5 𝑡𝑑
𝐾𝐶 = ( + ) ….. (11)
𝐾 𝑡𝑑 4 6𝜏
𝑡
6−2 𝑑
𝜏𝐷 = 𝑡𝑑 𝑡
𝜏
….. (12)
22+3 𝑑
𝜏
4) For Proportional-Integral-Derivative controller:
1 𝜏 4 𝑡𝑑
𝐾𝐶 = ( + ) ….. (13)
𝐾 𝑡𝑑 3 4𝜏
𝑡
32+6 𝑑
𝜏𝐼 = 𝑡𝑑 𝑡
𝜏
….. (14)
13+8 𝑑
𝜏
4
𝜏𝐷 = 𝑡𝑑 𝑡 ….. (15)
11+2 𝑑
𝜏

Process Control 47 Fourth Class

Dr. Zaidoon M. Shakor
 ym ym Actual
response
B B

Approximate
response
S Slop = S

td t td t

(a) (b)
Fig. (1) (a) Curve for Cohen-Coon tuning.
(b) Curve approximation with a first order dead-time system.

Example: consider the system with the following elements

1 1
𝐺𝑝 = 𝐺𝑚 = 𝐺𝑓 = 1
(5𝑠+1)(2𝑠+1) 10𝑠+1
Determine the optimum control elements using PRC method
1 1
𝐺𝑃𝑅𝐶 = 𝐺𝑓 𝐺𝑝 𝐺𝑚 = ∗ (10𝑠+1) ∗ 1
(5𝑠+1)(2𝑠+1)
Then draw Y(t) against t by using partial fraction of the above eqn.
𝐾𝑒 −𝑡𝑑𝑠
Then 𝐺𝑃𝑅𝐶 =
𝜏𝑠+1
From Fig. (1)
K=gain=B/A=1.0
S=slope at the inflection point=0.05
B=Ultimate response = 1.0
τ = effective time constant= B/S = 1/0.05 = 20
td = dead time =2.5
1.0𝑒 −2.5𝑠
∴ 𝐺𝑃𝑅𝐶 =
20𝑠 + 1
Using Cohen and Coon setting

For Proportional controller

1 𝜏 𝑡𝑑 1 20 2.5
𝐾𝑐 = (1 + ) = × (1 + ) = 8.33
𝐾 𝑡𝑑 3𝜏 1 2.5 3 × 20
For PI controller
1 𝜏 𝑑 𝑡
𝐾𝑐 = (0.9 + 12𝜏 ) = 7.3
𝐾 𝑡𝑑

Process Control 48 Fourth Class

Dr. Zaidoon M. Shakor
 30+3𝑡𝑑 /𝜏
𝜏𝐼 = 𝑡𝑑 = 6.6
9+20𝑡𝑑 /𝜏
For PID controller
1 𝜏 4 𝑡
𝐾𝑐 =
𝐾 𝑡𝑑
(3 + 4𝜏𝑑 ) = 10.9
32 + 6𝑡𝑑 /𝜏
𝜏𝐼 = 𝑡𝑑 = 5.85
13 + 8𝑡𝑑 /𝜏
4
𝜏𝐷 = 𝑡𝑑 = 0.89
11 + 2𝑡𝑑 /𝜏

Example: Consider the feedback control system for the stirred-tank blending process
shown in Fig. 1 and the following step test. The controller was placed in manual, and
then its output was suddenly changed from 30% to 43%. The resulting process reaction
curve is shown in Fig.(2). Thus, after the step change occurred at t = 0, the measured
exit composition changed from 35% to 55% (expressed as a percentage of the
measurement span), which is equivalent to the mole fraction changing from 0.10 to
0.30. Select the PID controller settings using the Cohen-Coon technique.

Figure (1) Composition control system for a stirred-tank blending process.

Process Control 49 Fourth Class

Dr. Zaidoon M. Shakor
 Figure (2) Process reaction curve.

Cohen and Coon Controller Settings

P PI PID
Kc 1 𝜏 𝑡𝑑 1 𝜏 𝑡𝑑 1 𝜏 4 𝑡𝑑
𝐾𝑐 = (1 + ) 𝐾𝑐 = (0.9 + ) 𝐾𝑐 = ( + )
𝐾 𝑡𝑑 3𝜏 𝐾 𝑡𝑑 12𝜏 𝐾 𝑡𝑑 3 4𝜏
τI - 30 + 3𝑡𝑑 /𝜏 32 + 6𝑡𝑑 /𝜏
𝜏𝐼 = 𝑡𝑑 ( ) 𝜏𝐼 = 𝑡𝑑 ( )
9 + 20𝑡𝑑 /𝜏 13 + 8𝑡𝑑 /𝜏
τD - - 4
𝜏𝐷 = 𝑡𝑑 ( )
11 + 2𝑡𝑑 /𝜏
 𝑡𝑑 = 𝑡𝑖𝑚𝑒 𝑑𝑒𝑙𝑎𝑦 (𝜃)

Answer
A first-order-plus-time-delay model can be developed from the process reaction curve
in Fig. (2) Using the graphical method. The tangent line through the inflection point
intersects the horizontal lines for the initial and final composition values at 1.07 min
and 7.00 min, respectively.
55 − 35
𝑆= = 3.3726 %/min
7 − 1.07
𝑂𝑢𝑡𝑝𝑢𝑡 𝐵 55 − 35 20
𝐾= = = = = 1.5384 %/%
𝐼𝑛𝑝𝑢𝑡 𝐴 43 − 30 13
∆𝑌
𝜏𝑑 = = 1.07 𝑚𝑖𝑛
∆𝑋
Process Control 50 Fourth Class
Dr. Zaidoon M. Shakor
 𝐵 20
𝜏= = = 5.93
𝑆 3.3726

The resulting empirical process model can be expressed as:

𝑋(𝑠) 1.538𝑒 −1.07
= 𝐺𝑝𝑟𝑐 =
𝑃(𝑠) 5.93s + 1
1 𝜏 4 𝑡
𝐾𝑐 =
𝐾 𝑡𝑑
(3 + 4𝜏𝑑 )= 4.967
32+6𝑡𝑑 /𝜏
𝜏𝐼 = 𝑡𝑑 ( )=2.45 min
13+8𝑡 /𝜏 𝑑
4
𝜏𝐷 = 𝑡𝑑 ( ) = 0.3767 𝑚𝑖𝑛
11 + 2𝑡𝑑 /𝜏

Example: Table (1) shows experimental process reaction curve of an open loop system
with a PI controller. Using the data in given in the table
(a) Approximate the open–loop response with that of a first–order system plus dead
time (i.e., find the values of K, τ, and td).
(b) Select the controller settings using the Cohen–Coon technique.

Table (1)
Time (min) -10 -5 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
Manipulated
10 10 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15
input
Measurement of
0.650 0.650 0.650 0.651 0.652 0.668 0.735 0.817 0.881 0.979 1.075 1.151 1.213 1.239 1.262 1.311 1.329 1.338 1.350 1.350 1.350
output

Cohen and Coon Controller Settings

P PI PID
1 𝜏 𝑡𝑑 1 𝜏 𝑡𝑑 1 𝜏 4 𝑡𝑑
Kc 𝐾𝑐 = (1 + ) 𝐾𝑐 = (0.9 + ) 𝐾𝑐 = ( + )
𝐾 𝑡𝑑 3𝜏 𝐾 𝑡𝑑 12𝜏 𝐾 𝑡𝑑 3 4𝜏
30 + 3𝑡𝑑 /𝜏 32 + 6𝑡𝑑 /𝜏
τI - 𝜏𝐼 = 𝑡𝑑 ( ) 𝜏𝐼 = 𝑡𝑑 ( )
9 + 20𝑡𝑑 /𝜏 13 + 8𝑡𝑑 /𝜏
4
τD - - 𝜏𝐷 = 𝑡𝑑 ( )
11 + 2𝑡𝑑 /𝜏

Answer

Process Control 51 Fourth Class

Dr. Zaidoon M. Shakor
 1.4

1.3

1.2

1.1
Y(t)
1

0.9

0.8

0.7

0.6
0 20 40 60 80 100
Time (min)

Ke  t d s
Then G PRC 
s1
From Fig(1)
K=gain=B/A=(1.35-0.65)/(15-10)=0.7/5=0.14
S=slope at the inflection point=(1.35-0.65)/(60-14)= 0.0152
B=Ultimate response = 0.7
τ = effective time constant= B/S = 0.7/0.0152= 46
td = dead time =14
0.14e 14 s
 GPRC 
46s1
Using Cohen and coon setting
For PI controller
1 t 
Kc   0.9 d   21.71
K td  12 
303t d / 
 I t d  28.68
9  20t d / 

2) Ziegler-Nichols Method
The method of Ziegler and Nichols known as the continuous cycling method. In this
method, integration and derivative terms of the controller are disabled and the
proportional gain is increased until a continuous oscillation occurs at ultimate gain Kcu
for the closed loop system. Considering Kcu and its related oscillating ultimate period,
Pu , the controller parameters can be calculated from the equation bellow.
𝐾𝑐𝑢 ⇒ 𝑓𝑟𝑜𝑚 𝑡ℎ𝑒 𝑝𝑟𝑜𝑐𝑒𝑑𝑢𝑟𝑒 𝑜𝑓 𝑡ℎ𝑒 continuous cycling method

Process Control 52 Fourth Class

Dr. Zaidoon M. Shakor
 2𝜋
𝑃𝑢 = 𝑤ℎ𝑒𝑛 (𝑤) = 𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦
𝑤

Pu Pu

Pu

1) For Proportional controller:

𝐾𝑐𝑢
𝐾𝑐 = ….. (11)
2
2) For Proportional-Integral controller:
𝐾𝑐𝑢
𝐾𝑐 = ….. (12)
2.2
𝑃𝑢
𝜏𝐼 = ….. (13)
1.2
3) For Proportional-Integral-Derivative controller:
𝐾𝑐𝑢
𝐾𝑐 = ….. (14)
1.7
𝑃𝑢
𝜏𝐼 = ….. (15)
2
𝑃𝑢
𝜏𝐷 = ….. (16)
8

Also, Ziegler-Nichols used another method bode diagram of two graphs: one is a plot
of the logarithm of the magnitude of sinusoidal transfer function; the other is a plot of
phase angle; both are plotted against the frequency on a logarithm scale as shown in
Fig. (2).
Gain margin (GM) and crossover frequency (𝜔) can be found from two plots therefore,
the ultimate gain and period of oscillation are calculated from following:

K u
 20 log GM  ….. (17)

P 
23.1428 ….. (18)
u


Process Control 53 Fourth Class

Dr. Zaidoon M. Shakor
 1.0

Gain
A.R
margin

0o

Ø(I)
Phase
Ø
margin
-180o

ωCO
ω

Fig. (2): Definition of gain and phase margins.

Example: Determine the Ziegler – Nichols tuning parameters for a PID controller, with
the given transfer functions. Assume that the time constants have units of minutes.
5 0.4 2
𝐺𝑣 = 𝐺𝑚 = 𝐺𝑝 =
2𝑆+1 5𝑆+1 𝑆+1
Answer
Continuous cycling method
1+ GOL = 0
1+Gv Gm Gp Gc = 0
Put Gc = Kcu
5 0.4 2
1+( ∗ ∗ ∗ 𝐾𝑐𝑢 ) = 0
2𝑆+1 5𝑆+1 𝑆+1
4 𝐾𝑐𝑢
1+( )=0
(2𝑆 + 1)(5𝑆 + 1)(𝑆 + 1)
10 S3 + 17 S2 + 8 S +1 + 4 Kcu = 0
Put S=iw , i= -1
10 (iw)3 + 17 (iw)2 + 8 (iw) +1 + 4 Kcu = 0
-10 iw3 – 17 w2 + 8 iw +1 + 4 Kcu = 0

Process Control 54 Fourth Class

Dr. Zaidoon M. Shakor
Real : – 17 w2 +1 + 4 Kcu = 0
Imaginary: -10 iw3 + 8 iw = 0
10 iw3 = 8 iw
w2 = 0.8 w = 0.894
-17 * 0.8 + 1 + 4 Kcu = 0
Kcu = 3.15
2𝜋 2𝜋
𝑃𝑢 = = = 7.03 𝑚𝑖𝑛
𝑤 0.894
For Kc , ꚍI , ꚍD
𝐾𝑐𝑢 3.15
𝐾𝑐 = = = 1.85
1.7 1.7
𝑃𝑢 7.03
𝜏𝐼 = = = 3.52 𝑚𝑖𝑛
2 2
𝑃𝑢 7.03
𝜏𝐷 = = = 0.879 𝑚𝑖𝑛
8 8

Process Control 55 Fourth Class

Dr. Zaidoon M. Shakor